AI Paper Atlas

Research map / Adversarial robustness and security / Certified and empirical robustness

Defending Against Image Corruptions Through Adversarial Augmentations

Dan Andrei Calian, Florian Stimberg, Olivia Wiles, Sylvestre-Alvise Rebuffi, András György, Timothy A Mann, Sven Gowal

ICLR 2022 · Adversarial robustness and security › Certified and empirical robustness · approach: Augmentation and diffusion purification · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Add AdA (EDSR) to AugMix + DeepAugmentCIFAR-10 classifier · CIFAR-10-CmCE: 7.99 -> 7.83Improved
Add AugMix to AdA (EDSR) (ablation)CIFAR-10 classifier · CIFAR-10-CmCE: 15.47 -> 9.40Improved
Add DeepAugment to AdA (EDSR) + AugMix (ablation)CIFAR-10 classifier · CIFAR-10-CmCE: 9.40 -> 7.83Improved
Use AdA (All) with AugMix and DeepAugment instead of nominal training with bothImageNet classifier · ImageNet-CmCE: 70.05 -> 62.90Improved
Use EDSR rather than CAE for AdA (ablation)ImageNet classifier · ImageNet-Crobust accuracy, L2 norm-bounded perturbations: 13.76 -> 32.93Improved
Use VQ-VAE or CAE as the AdA corruption network (ablation)CIFAR-10 classifier · CIFAR-10-CmCE: 26.77; 29.15Hurt
Combine AdA with DeepAugment (ablation)CIFAR-10 classifier · CIFAR-10.1accuracy: 90.35 -> 82.20Hurt
Combine AdA (All) with AugMix, DeepAugment, or both (ablation)ImageNet classifier · ImageNet-v2accuracy: 61.30; 59.43; 59.06; 57.57Hurt
Add AugMix to AdA (ablation)CIFAR-10 classifier · L2 norm-bounded perturbationsrobust accuracy: AdA: 9.28; 9.76; AdA + AugMix: 6.90; 6.09Mixed
Add DeepAugment after combining AdA with AugMix (ablation)CIFAR-10 classifier · L-infinity norm-bounded perturbationsrobust accuracy: 35.13 -> 49.95; 29.50 -> 34.82Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy